agentleFS
bitwize-music-studiobitwize-music-studio/claude-ai-music-skills/skills/suno-voice/SKILL.md

suno-voice

Skill668 starsChanged 10 months agoCC0-1.0View on GitHub

Puts one of the user's saved Suno Voices on a track or a whole album so /bitwize-music:generate sings in it, takes it off again, lists the Voices on their Suno account, or gives them Suno's link to make a new one. Checks the Voice exists first; nothing is generated or charged. Use when the user says "use my <name> voice", "sing this album in <Voice>", "what voices do I have", "take the voice off track 3" or "make a new voice".

The file

We split this skill into the rules it gives an agent

5 rules. About 0 are common, 2 are distinctive, and 3 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈2

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 3

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Follow § Preflight in the reference doc: find voices-list, voices-start and account-usage on the connected Suno server, load their current …

    line 36
  2. Distinctive:

    Not connected → say a Voice is checked against the Suno account through Suno's MCP server, give the setup command from § Preflight, and …

    line 38
  3. Distinctive:

    Only when the user asks for one.

    line 78
  4. Very specific:

    Call voices-start. It takes no arguments.

    line 80
  5. Very specific:

    Never ask the user for recordings or audio files: Suno takes them on that page.

    line 89

≈ “Common” is an estimate: at least 1 in 5 of the 372,429 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Your Task

Input: $ARGUMENTS

Work with the user's saved Suno Voices: list them, put one on a track or an album, take one off, or hand over Suno's link for making a new one.

Nothing here generates or costs anything. The Voice sings at the next /bitwize-music:generate of the track, at that command's usual cost.


Suno Voice Skill

A track sings in a saved Voice when the Voice row of its ## Cover / Voice Setup block names it (templates/track.md § Voice Selection). /bitwize-music:generate reads that row and finds the Voice's id on the user's Suno account. Background: ${CLAUDE_PLUGIN_ROOT}/reference/suno/mcp-api-mapping.md § Covers and Voices.

Step 1: Resolve

  • list (or "what voices do I have") → Step 3.
  • create (or "make a new voice") → Step 6.
  • <album> [<track>] "<Voice>" → Step 4. No track means every track of the album that has lyrics.
  • <album> [<track>] --off (or "take the voice off") → Step 5.

Step 2: Check the Suno Server

Follow § Preflight in the reference doc: find voices-list, voices-start and account-usage on the connected Suno server, load their current definitions, and read account-usage once.

  • Not connected → say a Voice is checked against the Suno account through Suno's MCP server, give the setup command from § Preflight, and stop.
  • features lacks voices → say Suno's Voices need a Pro or Premier plan, and stop.

Taking a Voice off (Step 5) needs none of this.

Step 3: List the Voices

voices-list (the user's own Voices, newest first). Show the names on one line each, no more than 20:

Your Voices on Suno (newest first):
  RTB · Missed Connections · ASCII · Mac · TayTay · Player Two

next_cursor set → say there are more and offer the next page. Voices they loved on suno.com are in collection: "loved"; show those when asked.

Step 4: Put a Voice on a Track or an Album

  1. Find it. voices-list(query: <name>). Nothing → once more with collection: "loved".
  2. Choose. One match, or one whose name is exactly the one asked for (ignoring case) → that Voice. Several otherwise → show their names and ask which. None → say it is not on their Suno account and show their Voices (Step 3) so they can pick one. They can also make one with /bitwize-music:suno-voice create — say so, but do not call voices-start unless they ask.
  3. Write it. set_track_voice(album_slug, voice, track_slug, saved_from, vocal_identity):
    • voice: the Voice's name. Use its id instead when another of their Voices has the same name, or when the Voice is a public one found by id, so /generate finds exactly this Voice.
    • saved_from: https://suno.com/song/<root_clip_id> when voices-list gave a root_clip_id.
    • vocal_identity: the Voice's description when it has one.
    • track_slug: empty for the whole album.
  4. Report.
    Track 03 now sings in RTB. The next /bitwize-music:generate uses it
    (10 credits for two takes, as usual). Takes made earlier keep their voice.
    For an album, name the tracks it went on, and the instrumental tracks it skipped. An error (an instrumental track, an unknown track) → show it.

A Voice carries its own gender, so the track's Vocal Gender is not sent with it. A Voice made on an older Suno model may need "Upgrade Voice to v6" on suno.com before it sings well.

Step 5: Take a Voice Off

set_track_voice(album_slug, "—", track_slug). The track then sings in the voice its Style Box describes. Say which tracks changed; unchanged ones had no Voice.

Step 6: Make a New Voice

Only when the user asks for one.

  1. Call voices-start. It takes no arguments.
  2. A url comes back → give it to the user:
    Make your Voice on suno.com: <url>
    Record and verify it there. When it's ready, tell me its name and the
    track or album to use it on.
  3. No url → say what Suno answered and stop. An account signed in with an API key gets a 409; Voices are made by a user signed in on suno.com.

Never ask the user for recordings or audio files: Suno takes them on that page. Once the Voice is ready it appears in voices-list, and Step 4 puts it on a track.

Remember

  1. Check before writing — a Voice goes in a track file only after voices-list has found it.
  2. Free — nothing here generates or charges. Singing happens at /generate, with its own cost and go-ahead.
  3. The user names the Voice — never pick one for them; with several matches, ask.
  4. voices-start only on request — when a Voice is missing, offer it; call it only when they want to make one.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.